Francisco Ruben Castillo Soria

Deep learning (DL) has emerged as a promising alternative for signal detection in multiple-input multiple-output (MIMO) wireless communication systems, achieving a performance close to the optimal maximum likelihood (ML) criterion. However, implementing DL algorithms for MIMO signal detection still demands a high computational cost. Recently, a DL-based architecture with reduced complexity was proposed for MIMO signal detection. More specifically, a preprocessing stage was added to label the input signals, resulting in a more efficient detection scheme. However, even with this reduction in complexity, the proposed detection schemes remain computationally demanding, and this cost grows sharply as the system scales to higher-order modulations and/or large antenna arrays. To further reduce the complexity of these schemes, this paper proposes a graphics processing unit (GPU)-accelerated procedure for DL-based MIMO signal detection. The algorithm is accelerated by using two strategies in a CUDA-enabled PyTorch implementation. First, we optimize the training, which tightens the bit error rate (BER) gap from 1.00 dB to approximately 0.20 dB for the One-Hot (OH) labeling and from 2.00 dB to approximately 0.90 dB for the One-Hot-per-Antenna (OHA) labeling. Additionally, we optimize the Monte Carlo simulation using seven strategies, reducing the end-to-end evaluation time by 34.7% on an NVIDIA RTX 4090 GPU. These results aim to support future projects in evaluating DL-based MIMO systems using shorter computing windows.

LoRaWAN has emerged as an efficient solution for large-scale Internet of Things (IoT) deployments, where reliability, scalability, and energy efficiency are key performance indicators. Because experimental evaluation is costly, hard to replicate, and highly dependent on environmental conditions, FLoRa, an OMNeT++-based framework, has been widely adopted for LoRaWANperformance evaluation. However, we identified implementation-level inconsistencies in FLoRa that compromise simulation fidelity. The default channel model uses a fixed reference-loss anchoring at distance d₀, introducing systematic bias in received-power estimation having an adverse effect on the Adaptive Data Rate (ADR). In addition, the end-device ADR recovery behavior does not fully comply with the standard, and packet decapsulation is not correctly handled during the RX2 reception window. This work corrects these issues and reassesses network-level performance in urban and suburban deployments. In urban scenarios, the mean packet delivery rate (PDR) increases from 36.25% to 98%, while the mean energy per successful transmission decreases from 302 mJ to 100 mJ. In suburban scenarios, the mean PDR increases from 27.45% to 96%, while the mean energy per successful transmission decreases from 390 mJ to 102.2 mJ. In addition, ADR-related MAC commands per end device decrease significantly, reflecting an ADR behavior according to expected.

Reconfigurable intelligent surface (RIS)-assisted non-orthogonal multiple access (NOMA) has emerged as a promising technique to enhance spectral efficiency and coverage in fifth- and sixth-generation wireless networks. However, asymmetric indoor propagation conditions characterized by heterogeneous line-of-sight (LoS) and non-line-of-sight (NLoS) links often degrade user fairness. This paper investigates a downlink RIS-assisted NOMA system under the standardized 3GPP indoor office (InH) channel model to address fairness-oriented design under realistic link-budget constraints. We formulate an optimization problem for max–min fairness that jointly considers discrete RIS element partitioning and NOMA power allocation to achieve a symmetrical allocation of quality of service (QoS). To enable efficient computation, the non-convex problem is transformed into an epigraph form and solved using a low-complexity, bisection-based quasi-convex optimization framework combined with enumeration over RIS partitions. Numerical results demonstrate significant fairness gains; for instance, doubling the RIS array size yields a substantial improvement in the ergodic max–min rate, corresponding to approximately a 66% gain at moderate transmit power levels. Furthermore, by accounting for practical impairments such as imperfect successive interference cancellation (iSIC), imperfect channel state information (iCSI), and RIS implementation losses, the results reveal that fairness-optimal operation consistently prioritizes the far user to overcome severe indoor NLoS attenuation. The proposed framework is also compared with alternating optimization (AO)-based RIS-NOMA, conventional RIS beamforming without partition and RIS-assisted orthogonal multiple access (OMA) schemes. Simulation results confirm that the proposed framework achieves low computational complexity, making it suitable for practical indoor wireless environments.

This study investigates the impact of audio feedback on cognitive performance during VR puzzle games using EEG analysis. Thirty participants played three different VR puzzle games under two conditions (with and without audio) while their brain activity was recorded. To analyze concentration levels and neural engagement patterns, we employed spectral analysis combined with a preprocessing algorithm and an optimized Deep Neural Network (DNN) model. The proposed processing stage integrates feature normalization, automatic labeling based on Principal Component Analysis (PCA), and Gamma band feature extraction, transforming concentration detection into a supervised classification problem. Experimental validation was conducted under the two gaming conditions in order to evaluate the impact of multisensory stimulation on model performance. The results show that the proposed approach significantly outperforms traditional machine learning classifiers (SVM, LR) and baseline deep learning models (DNN, DGCNN), achieving a 97% accuracy in the audio scenario and 83% without audio. These findings confirm that auditory stimulation reinforces neural coherence and improves the discriminability of EEG patterns, while the proposed method maintains a robust performance under less stimulating conditions.

Head-on vehicle collision prevention remains a critical challenge in autonomous and manual driving, particularly for complex vehicular scenarios where conventional sensors face line-of-sight limitations. In this work, we propose a novel fully centralized warning system platform using continuous waveform (CW) signals and Doppler signature analysis. We use a propagation model to analyze Doppler effects in vehicular communication systems, validated empirically across two distinct driving environments (high-speed highway and medium-speed rural road). Our platform was developed using general-purpose equipment to generate a data set of spectrograms computed with the received radio-frequency (RF) CW signals. Furthermore, machine learning models (SVM/KNN/Boosted Trees) applied to spectrogram features reduced via Principal Component Analysis are used to classify five different vehicular events related to head-on collision. Our system achieves up to 99% classification accuracy while demonstrating that Doppler signatures in communication signals can be used to extract high-quality information for safety-critical sensing. Our results show robust performance in both test scenarios, with high-precision for oncoming vehicles at different speeds. The system’s success in using CW RF signals for sensing, establishes a foundation for Integrated Sensing and Communication (ISAC) implementations.

New communication systems require high spectral and energy efficiencies to meet the growing demand for services in future networks. In this paper, an efficient multiple parallel reconfigurable intelligent surfaces (RIS)-assisted multiuser (MU) multiple input-multiple output (MIMO) double quadrature spatial modulation (DQSM) downlink transmission system is presented. In the transmitter, the proposed N-RIS-MU-MIMO-DQSM system uses a modified block diagonalization technique and a genetic algorithm (GA) to jointly design the precoding signals required at the base station (BS) and the optimal phase changes required at multiple RISs. A reduced detection complexity and improved bit error rate (BER) performance are achieved by incorporating spatial modulation. The proposed system is compared under the same conditions and parameters with two reference systems, considering blind and optimized RISs approaches over correlated Rayleigh fading channels. Results show that compared with a similar system that does not use RISs, the proposed system has up to 30 dB gain in BER performance. Compared with a similar system based on conventional quadrature amplitude modulation (QAM), the proposed system has gains of up to 2–3 dB in BER performance and up to 55.8% lower detection complexity for the analyzed cases.

This paper investigates the application of neural networks for Quadrature Phase Shift Keying (QPSK) signal detection in Orthogonal Frequency Division Multiplexing (OFDM) systems designed for wideband vehicle-to-vehicle (V2V) communication environments. In such scenarios, inter-carrier interference (ICI) induced by doubly dispersive channels leads to a loss of subcarrier orthogonality, significantly impairing detection performance. Traditional linear and nonlinear ICI mitigation techniques exhibit fundamental trade-offs between bit error rate (BER) performance and computational complexity, limiting their effectiveness in highly dynamic V2V conditions. A low-complexity neural network is proposed for M-PSK detection under ICI. The approach includes a preprocessing step that converts interference mitigation into a classification task. Results show that this approach performs better than conventional non-linear algorithms while keeping complexity near to linear least squares detection.

Reconfigurable intelligent surfaces (RIS) and massive multiple input and multiple output (M-MIMO) are the two major enabling technologies for next-generation networks, capable of providing spectral efficiency (SE), energy efficiency (EE), array gain, spatial multiplexing, and reliability. This work introduces an RIS-assisted millimeter wave (mmWave) M-MIMO system to harvest the advantages of RIS and mmWave M-MIMO systems that are required for beyond fifth-generation (B5G) systems. The performance of the proposed system is evaluated under 3GPP TR 38.901 V16.1.0 5G channel models. Specifically, we considered indoor hotspot (InH)—indoor office and urban microcellular (UMi)—street canyon channel environments for 28 GHz and 73 GHz mmWave frequencies. Using the SimRIS channel simulator, the channel matrices were generated for the required number of realizations. Monte Carlo simulations were executed extensively to evaluate the proposed system’s average bit error rate (ABER) and sum rate performances, and it was observed that increasing the number of transmit antennas from 4 to 64 resulted in a better performance gain of ∼10 dB for both InH—indoor office and UMi—street canyon channel environments. The improvement of the number of RIS elements from 64 to 1024 resulted in ∼7 dB performance gain. It was also observed that ABER performance at 28 GHz was better compared to 73 GHz by at least ∼5 dB for the considered channels. The impact of finite resolution RIS on the considered 5G channel models was also evaluated. ABER performance degraded for 2-bit finite resolution RIS compared to ideal infinite resolution RIS by ∼6 dB.